Turn text into embeddings or search documents by meaning. OpenAI-compatible API with zero per-token cost.
Send text, get back numbers that capture its meaning. Use them for similarity, recommendations, or feed them into your own systems.
Store documents and search with natural language. Ask "how to deploy?" and find docs about "shipping to production".
Change your base URL and key — done. Works with official Python and JavaScript SDKs. No code changes needed.
Projects hold collections, collections hold documents. Each user's data is fully isolated. Scale to millions of vectors.
Any text, any language
1024 numbers capturing meaning
Or use the vectors yourself
| OpenAI | Embedding AI | |
|---|---|---|
| Cost | $0.02 / 1M tokens | Free |
| Model quality | MTEB: 62.3 | MTEB: 63.0 (BGE-M3) |
| Dimensions | 1536 | 1024 |
| Rate limits | Yes | None |
| Data privacy | Sent to OpenAI | Stays on your server |
| Vector storage | No | Built-in (Qdrant) |
| SDK compatible | — | Yes, drop-in |
Ready to start embedding?
Read the Docs →